A Digital Microfluidic Approach to Homogeneous Enzyme Assays

Jan 26, 2008 - Mathias Girault, Thomas Beneyton, Deniz Pekin, Lionel Buisson, Sabrina ... Jessica Olofsson, Shijun Xu, Gavin D. M. Jeffries, Aldo Jeso...
0 downloads 0 Views 519KB Size
Anal. Chem. 2008, 80, 1614-1619

A Digital Microfluidic Approach to Homogeneous Enzyme Assays Elizabeth M. Miller and Aaron R. Wheeler*

Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, ON, M5S 3H6, Canada, and Donnelly Centre for Cellular and Biomolecular Research, 160 College Street., Toronto, ON, M5S 3E1, Canada

A digital microfluidic device was applied to a variety of enzymatic analyses. The digital approach to microfluidics manipulates samples and reagents in the form of discrete droplets, as opposed to the streams of fluid used in channel microfluidics. This approach is more easily reconfigured than a channel device, and the flexibility of these devices makes them suitable for a wide variety of applications. Alkaline phosphatase was chosen as a model enzyme and used to convert fluorescein diphosphate into fluorescein. Droplets of alkaline phosphatase and fluorescein diphosphate were merged and mixed on the device, resulting in a 140-nL, stopped-flow reaction chamber in which the fluorescent product was detected by a fluorescence plate reader. Substrate quantitation was achieved with a linear range of 2 orders of magnitude and a detection limit of ∼7.0 × 10-20 mol. Addition of a small amount of a nonionic surfactant to the reaction buffer was shown to reduce the adsorption of enzyme to the device surface and extend the lifetime of the device without affecting the enzyme activity. Analyses of the enzyme kinetics and the effects of inhibition with inorganic phosphate were performed, and Km and kcat values of 1.35 µM and 120 s-1, respectively, agreed with those obtained in a conventional 384-well plate under the same conditions (1.85 µM and 155 s-1). A phototype device was also developed to perform multiplexed enzyme analyses. It was concluded that the digital microfluidic format is able to perform detailed and reproducible assays of substrate concentrations and enzyme activity in much smaller reaction volumes and with higher sensitivity than conventional methods. Enzymatic assays are a key component in a variety of applications, such as glucose monitoring in serum1 and screening for phenols2 or organophosphates3 in drinking water. Relative to other analytical techniques such as separations and mass spectrometry, enzyme assays can be used to detect or quantify the concentrations of small molecules in a format that is far more compact, portable, and easy-to-use. In addition, enzyme assays * To whom correspondence should be addressed. E-mail: awheeler@ chem.utoronto.ca. Tel: (416) 946 3864. Fax: (416) 946 3865. (1) Reach, G.; Wilson, G. S. Anal. Chem. 1992, 64, 381A-386A. (2) Chen, W. J. Anal. Chim. Acta 1995, 312, 39-44. (3) Mulchandani, A.; Chen, W.; Mulchandani, P.; Wang, J.; Rogers, K. R. Biosens. Bioelectron. 2001, 16, 225-230.

1614 Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

are uniquely suited for analyzing the activity of proteins (instead of merely detecting them). A challenge for enzyme assays is in analytical capacity and sensitivitysseveral different analyses may be required to characterize a highly complex sample (e.g., a cellular extract), and many replicates of each analysis must be performed to ensure the reproducibility of the data. As the number of analyses being performed increases, the cost of reagents and the amount of sample required (often available in limited quantities) become important factors. The technology of microfluidics, which can be used to examine many samples in parallel while consuming smaller amounts of samples and reagents than is required for conventional-scale methods, has the potential to make enzyme assays indispensable for a wide range of applications.4 Most microfluidic enzyme assays have been implemented via the channel microfluidic format, in which reagents and samples are transported through a system of enclosed micrometerdimension channels as streams of fluid. Hadd et al.5,6 performed some of the first enzyme assays in microchannels in a continuousflow format and demonstrated a 4 order of magnitude reduction in enzyme and substrate consumption over conventional methods. However, this system relied on diffusional mixing, and reaction times of up to 20 min were required to obtain accurate kinetic data. With the help of a microfabricated mixer, Burke et al.7 conducted a stopped-flow enzyme assay that consumed only 6 nL of enzyme and required less than 60 s to complete. While this approach greatly reduces reagent consumption and analysis times over macroscale methods, fabrication methods for such systems are often expensive and time-consuming.8 Once a particular arrangement of channels and reaction chambers is designed, a great deal of time and expense is required to change the device to suit a different application. The fabrication of multiplexed devices becomes a daunting prospect when features such as mixers must be incorporated, as even the simplest mixing schemes require an integrated network of many smaller features such as channels9 or raised ridges.10 (4) Miller, E. M.; Freire, S. L. S.; Wheeler, A. R. In Encyclopedia of Micro- and Nanofluidics; Li, D. Q., Ed.; Springer: Heidelberg, Germany, in press. (5) Hadd, A. G.; Raymond, D. E.; Halliwell, J. W.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 1997, 69, 3407-3412. (6) Hadd, A. G.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 1999, 71, 52065212. (7) Burke, B. J.; Regnier, F. E. Anal. Chem. 2003, 75, 1786-1791. (8) Dittrich, P. S.; Tachikawa, K.; Manz, A. Anal. Chem. 2006, 78, 3887-3908. (9) He, B.; Burke, B. J.; Zhang, X.; Zhang, R.; Regnier, F. E. Anal. Chem. 2001, 73, 1942-1947. 10.1021/ac702269d CCC: $40.75

© 2008 American Chemical Society Published on Web 01/26/2008

An alternative paradigm for microfluidics has recently emerged, in which small volumes of fluid are manipulated in the form of droplets on an open platform instead of as streams in enclosed channels. This approach, called digital microfluidics (DMF), exploits the electrowetting and dielectrophoresis forces generated when an electrical potential is applied to an electrode in an array.11-14 By applying electrical potentials to sequential electrodes, a droplet of fluid can be dispensed from a reservoir, transported to any position on the array, and merged with other droplets to perform reactions.15,16 This technique has been used to actuate a wide range of volumes (nL to µL), and unlike channel devices, there is no sample wasted in creating a small plug for analysis. In addition, each droplet is isolated from its surroundings rather than being embedded in a stream of fluidsa simple method of forming a microreactor in which there is no possibility that products will diffuse away. Perhaps most importantly, the geometry of the array ensures that any device design is reconfigurablessince a greater number of paths through the device are possible, a variety of functions can be performed without redesigning the device. There is currently much enthusiasm for applying DMF to multiplexed assays;17 however, it has only been applied to a few biological applications.18-25 One reason for the lack of applications for this technique is that fabrication of DMF devices is arduous and timeconsuming; we recently developed a suite of rapid prototyping methods for DMF devices,26-28 which should make the method available to all who wish to use it. A wide variety of channel microfluidic devices has been developed for enzyme studies,5-7,29 but the potential of digital devices has been largely unexplored. Taniguchi et al.18 demonstrated the movement and merging of droplets of luciferase and luciferin on an electrowetting-based device, but this system required a hydrophobic liquid (oil) medium surrounding the (10) Stroock, A. D.; Dertinger, S. K. W.; Adjari, A.; Mezic, I.; Stone, H. A.; Whitesides, G. M. Science 2002, 295, 647-651. (11) Lee, J.; Moon, H.; Fowler, J.; Schoellhammer, T.; Kim, C.-J. Sens. Actuators, A 2002, 95, 259-268. (12) Pollack, M. G.; Fair, R. B.; Shenderov, A. D. Appl. Phys. Lett. 2000, 77, 1725-1726. (13) Washizu, M. IEEE Trans. Ind. Appl. 1998, 34, 732-737. (14) Velev, O. D.; Prevo, B. G.; Bhatt, K. H. Nature 2003, 426, 515-516. (15) Fowler, J.; Moon, H.; Kim, C.-J. Proc. IEEE Conf. Microelectromech. Syst. 2002, 97-100. (16) Cho, S. K.; Moon, H.; Kim, C.-J. J. Microelectromech. Syst. 2003, 12, 7080. (17) Mukhopadhyan, R. Anal. Chem. 2006, 78, 1401-1404. (18) Taniguchi, T.; Torii, T.; Higuchi, T. Lab Chip 2002, 2, 19-23. (19) Srinivasan, V.; Pamula, V. K.; Fair, R. B. Lab Chip 2004, 4, 310-315. (20) Srinivasan, V.; Pamula, V. K.; Fair, R. B. Anal. Chim. Acta 2004, 507, 145150. (21) Jary, D.; Chollat-Namy, A.; Fouillet, Y.; Boutet, J.; Chabrol, C.; Castellan, G.; Gasparutto, D.; Peponnet, C. In NSTI Nanotech Technical Proceedings 2006, Vol. 2, pp 554-557. (22) Chang, Y. H.; Lee, G. B.; Huang, F. C.; Chen, Y. Y.; Lin, J. L. Biomedical Microdevices 2006, 8, 215-225. (23) Wheeler, A. R.; Moon, H.; Kim, C.-J.; Loo, J. A.; Garrell, R. L. Anal. Chem. 2004, 76, 4833-4838. (24) Wheeler, A. R.; Moon, H.; Bird, C. A.; Loo, R. R. O.; Kim, C.-J.; Loo, J. A.; Garrell, R. L. Anal. Chem. 2005, 77, 534-540. (25) Moon, H.; Wheeler, A. R.; Garrell, R. L.; Loo, J. A.; Kim, C.-J. Lab Chip 2006, 6, 1213-1219. (26) Watson, M.; Abdelgawad, M.; Ye, G.; Yonson, N.; Trottier, J.; Wheeler, A. R. Anal. Chem. 2006, 78, 7877-7885. (27) Abdelgawad, M.; Wheeler, A. R. Adv. Mater. 2007, 19, 133-137. (28) Abdelgawad, M.; Wheeler, A. R. Microfluidics and Nanofluidics; in press. DOI: 10.1007/s10404-007-0190-3. (29) Duffy, D. C.; Gillis, H. L.; Lin, J.; Sheppard, N. F. Jr.; Kellogg, G. J. Anal. Chem. 1999, 71, 4669-4678.

droplet, and no quantitative data were obtained. A two-enzyme assay for the detection of glucose was reported by Srinivasan et al.19, 20 that also required a silicone oil medium to achieve droplet movement and prevent analyte adsorption to the device surface. While this method proved to be effective for the quantitation of glucose in a variety of biological fluids, a detailed analysis of enzyme activity was not achieved. It would be desirable to develop an assay that does not require oil, as it can make working with some solvents such as ethanol or methanol impossible; it is also likely that hydrophobic analytes may partition out of the aqueous droplet and into the oil, resulting in incomplete reactions, inaccurate quantitation, and cross-contamination. Jary et al.21 developed a microfluidic microprocessor for DNA repair enzyme analysis that was just as sensitive as macroscale methods, but with a more limited dynamic range. An ideal microfluidic assay system would have high sensitivity, yet retain the flexibility and dynamic range of its large-scale counterpart. Here, we demonstrate the first application of digital microfluidic methods to the detection and quantification of small molecules via an enzyme assay, as well as for the study of enzyme kinetics, with no need for liquid suspending media (e.g, silicone oil). We show that the DMF assay has better sensitivity than macroscale methods, but without sacrificing dynamic range. This technique has great potential as a simple yet versatile analytical tool for implementing enzymatic analyses on the microscale. EXPERIMENTAL SECTION Reagents and Materials. Alkaline phosphatase (type VII-L from bovine intestinal mucosa), diethanolamine, magnesium chloride, Fluorinert FC-40, sodium phosphate dibasic, and Pluronic F-127 were purchased from Sigma Chemical (Oakville, ON, Canada). Fluorescein diphosphate (FDP) was purchased from Molecular Probes (Invitrogen Canada, Burlington, ON, Camada). Parylene-C dimer was from Specialty Coating Systems (Indianapolis, IN), and Teflon-AF was purchased from DuPont (Wilmington, DE). Clean room reagents and supplies included Shipley S1811 photoresist and developer from Rohm and Haas (Marlborough, MA), solid chromium and gold from Kurt J. Lesker Canada (Toronto, ON, Canada), CR-4 chromium etchant from Cyantek (Fremont, CA), hexamethyldisilazane (HMDS) from Shin-Etsu MicroSi (Phoenix, AZ), AZ-300T photoresist stripper from AZ Electronic Materials (Somerville, NJ), and conc. sulfuric acid and hydrogen peroxide (30%) from Fisher Scientific Canada (Ottawa, ON, Canada). Piranha solution was prepared as a 3:1 (v/v) mixture of sulfuric acid and hydrogen peroxide. Device Fabrication. Digital microfluidic devices were formed using conventional methods in the University of Toronto Emerging Communications Technology Institute (ECTI) fabrication facility. Glass wafers were cleaned in piranha solution (10 min) and then coated with chromium (10 nm) and gold (100 nm) by electron beam deposition. After rinsing and baking on a hot plate (115 °C, 5 min), the substrates were primed by spin-coating with HMDS (3000 rpm, 30 s) and then spin-coated with Shipley S1811 photoresist (3000 rpm, 30 s). Substrates were prebaked on a hot plate (100 °C, 2 min) and exposed through a photomask using a Suss Mikrotek mask aligner. Substrates were developed in MF321 (3 min) and then postbaked on a hot plate (100 °C, 1 min). After photolithography, substrates were immersed in gold etchant Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

1615

Figure 1. Fluorescent enzymatic assay on a DMF device. A droplet containing FDP was dispensed from the reservoir on the right (a) and a droplet of AP was dispensed from the reservoir on the left (a, b). When the droplets were merged under fluorescent illumination, the product was observed at the interface of the droplets (c). After active mixing, the reaction proceeded to completion (d). Lines have been added to the images in (a, b) to indicate the location and movement of droplets.

(60 s) followed by chromium etchant (60 s). Finally, the remaining photoresist was stripped in AZ300T (5 min) in an ultrasonic cleaner. After forming electrodes, devices were coated with 2 µm of parylene C and 50 nm of Teflon-AF. Parylene C was applied using a vapor deposition instrument (Specialty Coating Systems), and Teflon-AF was spin-coated (1% w/w in Fluorinert FC-40, 2000 rpm, 60 s) and then postbaked on a hot plate (160 °C, 10 min). To actuate droplets, the polymer coatings were locally removed from the contact pads by gentle scraping with wafer tweezers. In addition to patterned devices, unpatterned indium-tin oxide (ITO)coated glass substrates (Delta Technologies Ltd, Stillwater, MN) were coated with Teflon-AF (50 nm, as above). Droplet Actuation. Each device was assembled with an unpatterned ITO/glass top plate and a patterned bottom plate separated by a spacer formed from one piece of double-sided tape (∼70 µm thick). The 140-nL droplets in devices with this arrangement had a diameter of ∼1.6 mm (∼40% of the size of a well in a 384-well plate). As described previously,11,16,24 droplets were sandwiched between the two plates and actuated by applying electric potentials between the top electrode and sequential electrodes on the bottom plate. Applied potentials (60-80 VRMS) were generated by amplifying the output of a function generator operating at 18 kHz. Driving potentials were applied manually to exposed contact pads on the bottom plate surface. Droplet actuation was monitored and recorded by a CCD camera mated to a stereomicroscope with fluorescence imaging capability (Olympus Canada, Markham, ON, Canada). All devices had 1 mm × 1 mm actuation electrodes, with an interelectrode gap of 4075 µm. 1616

Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

Analysis of Enzyme Activities and Kinetics. Alkaline phosphatase solutions were prepared in 10 mM diethanolamine (DEA) buffer, pH 10.1, containing 1 mM MgCl2 and 0.1% (w/v) Pluronic F-127. Enzyme concentrations were 6 U/mL for standard curve experiments and for observation through the microscope and 0.21 U/mL for kinetic analyses. Fluorescein diphosphate solutions of various concentrations were prepared in 10 mM DEA buffer. For enzyme inhibition experiments, fluorescein diphosphate solutions containing various concentrations of inorganic phosphate were also prepared. For all quantitative experiments, a digital microfluidic device was manually positioned on the top of a 384-well microtiter plate, and a 650-nL reservoir droplet of enzyme or substrate solution was pipetted onto each of two large electrodes (2 mm × 2 mm). As depicted in Figure 1, 70-nL droplets of each reagent were actively dispensed from the reservoirs (as described previously16). Briefly, two electrodes adjacent to the reservoir were actuated sequentially to form a finger of fluid extending onto the array; upon simultaneous actuation of the target electrode and the reservoir electrode; the finger was split, resulting in a dispensed droplet with a volume defined approximately by the electrode and spacer dimensions (1 mm × 1 mm × 70 µm). The two dispensed reagent droplets were then merged and mixed by moving the coalesced droplet around a loop of six actuation electrodes for 15 s. After mixing, the device (still positioned on the microtiter plate) was inserted into a PheraStar multiwell plate reader (BMG Labtech, Durham, NC) for fluorescence detection (λex ) 485 nm; λem ) 520 nm; focal height, 15.0 mm; gain, 1462 or 90 for substrate quantitation and kinetics experiments, respectively). After each assay, devices were rinsed in DEA buffer and DI water and allowed to dry. Because of the finite time required

Figure 3. Effect of Pluronic F-127 on enzyme assays. An APFDP reaction was allowed to proceed to completion in a well plate under three sets of conditions: (1) without any Pluronic added to the reaction mixture, (2) with Pluronic added to the wells after the reaction was completed, and (3) with Pluronic added to the wells at the start of the reaction. Error bars are (1 SD. While the presence of the polymer does reduce the fluorescent signal at the detector, the reduction is the same whether the Pluronic is added at the beginning of the reaction or at completion. The activity of the enzyme appears to be unaffected.

Figure 2. FDP calibration curves. On a 384-well plate (a) and a DMF device (b), the fluorescence response was linear over a range of 2 orders of magnitude of substrate concentration. Error bars are (1 SD. In both experiments, the error in the data points ranged from